Fix API key errors

Most chat errors come down to an API key that's missing, mistyped, or out of credit. Find your message below.

"{Provider} API Key not found"

For example "OpenAI API Key not found. Please set it in your profile settings."

There's no key for the provider of the model you picked.

1. Open Settings → API Keys.
2. Paste a key for that provider and click Securely Save Keys.
3. Or pick a model from a provider you already have a key for.

Remember that keys are stored per browser. On a new device or browser, add them again or restore them with Cloud sync.

"{Provider} API Key is incorrect"

The provider rejected the key.

• Copy the key again from the provider's dashboard - make sure there are no spaces before or after it.
• Check the key hasn't been revoked or expired.
• Some providers have separate keys for different products. Make sure it's an API key for the model API.

"Rate limit exceeded. Please try again later."

The provider is limiting how many requests you can send.

• Wait a minute and try again.
• Check your plan and usage limits in the provider's dashboard - new accounts often have low limits.
• Make sure your provider account has billing set up or credit available. Many providers return errors once prepaid credit runs out.

"Custom API Key not found / is incorrect"

This is about a custom model. Open it in the Models sidebar and check its API Key and Base URL.

Qwen errors

Qwen needs both an API key and your Model Studio base URL. Check that the base URL is the OpenAI-compatible endpoint for your region.

In a company workspace

If you see "API key set by admin", you can't change the key yourself. Ask your workspace admin to check the key under Settings → API Keys in the Business console.

Still stuck?

Contact support with the exact error message and the model you were using. Never send us your API key.